Dave's Free Press: Journal

violence, pornography, and rude words for the web generation


Recent posts


Recently commented posts


Journals what I read

geeky politics rant silly religion meta music perl weird drinking culture london language transport sport olympics hacking media maths web photography etiquette spam amazon film bastards books bryar holidays palm telecoms cars travel yapc bbc clothes rsnapshot phone whisky security home radio lolcats deafness environment curry art work privacy iphone linux bramble unix go business engineering kindle gps economics latin anglo-saxon money cars environment electronics
Tue, 1 Mar 2005

Some people should be certified

It is something of a tradition amongst groups of Perl programmers to argue about the merits or otherwise of a Perl certification programme similar to the very popular* Java certification programme. This week, it seems to be the turn of the Boston Perl Mongers. Normally, this wouldn't be something worth writing about, except for this gem of a posting to the list:

From: "John Redford" <[elided]@hotmail.com>

I am anti-certification. Why? To put it extremely bluntly: certifications are socialist.


In reality, certifications would be given to people who paid for them, regardless of what they know.

Only certifying those who pay for certification, rather than those who deserve it, or those who need it, is in fact capitalist. Nice one John.

The hiring practices of companies would require that the certification be held

The fact that university degrees exist, that I don't have one, and that I have a well-paid job as a professional programmer would seem to be something of a contra-indication. Now of course drawing any conclusions from such a small sample is foolish. However, I estimate (simply based on working with many hundred people over the past many years) that something like 20% of professional programmers have no degree at all, let alone one in computing, and that in fact under 30% have degrees in computing. Of those 20% who are degree-less, a goodly proportion have no qualifications in computing at all.

For a small bribe, the person administering the test would provide the candidate with the answers.

A "small bribe" would not be enough to compensate the tester for losing his job and his income. This is a very real risk of which trainers and testers of my acquaintance are all too aware. It is in the certificate issuer's interests to ensure that this doesn't happen, for they want their certificates to be seen as being worth something. I'd not be in the least bit surprised if they occasionally tried to "sting" testers and trainers.

Deciding not to hire a person who holds certifications becomes hard to justify.

I have recommended that my various employers not employ certain people who are laden down with certifications by saying things like "this guy doesn't have sufficient experience".

Firing a person for incompetence would become even more problematic, as their holding of a certification would be considered proof of their competence.

Not even the management types who seem to like asking for certifications really believe this! They have, quite simply, too much experience of certified people being shit. Firing a person for incompetence is in fact quite easy if you catch the incompetence early, regardless of what bits of paper they have. It can be quite hard if you don't catch it early, for reasons which are entirely different from what qualifications the incompetent employee has.

The job market moves, to a small or large degree, towards a static pool of incompetent criminals who cannot be fired and cannot contribute.

For this, I give John ten out of ten for use of rhetorical hyperbole, but zero out of ten for actually making sense.

* I originally wrote "successful" here but changed it because it really isn't successful, as can be seen by comparing the quality of yer average Java program to yer average Perl program. They both suck.

Posted at 19:58 by David Cantrell